Passivity-Based State of Charge Estimation for Switched Li-Ion Batteries with Modeling and Measurement Uncertainties

The objective of this paper is to introduce a model-based approach for state of charge (SOC) estimation tailored specifically by incorporating switched dynamics and passivity theory for the first time for lithium-ion batteries (LIBs). Revealing the battery's impact on relaxation and exploring the intricate correlation between SOC and EMF, we propose a switched battery model based on a linear equivalent circuit model (ECM), establishing a robust foundation for precise SOC estimation. The method for estimating SOC utilizes an observer with a passivity switching approach, ensuring resilience against uncertainties in batteries with various phenomena; as a result, a novel observer design is proposed. The switched systems and passivity theory synergy enhance SOC estimation precision, demonstrating adaptability across various electric vehicle (EV) battery types, with specific considerations for piecewise linearization. Treating uncertainties in ECM parameters as modeling errors constrained by passivity performance, our proposed method proves effective for diverse conditions encountered in EV applications. A simulation example demonstrates the practical utility and benefits of the proposed method, offering a resilient solution for accurate SOC estimation in LIBs for EVs.